Dry Eye
Conditions
Keywords
dry eye
Brief summary
The bulk of dry eye patients are found in the community. The lack of satisfactory protocols and confidence is a significant deterrent for practitioners to manage such patients, which may result in inaccurate referrals, and unhappy patients. Problems are compounded by comorbidities of dry eye, even if these are not diagnosed formally. Aligning with the healthcare strategy to move beyond healthcare to health, and beyond hospital care to community care, investigators propose that the confidence of primary carers be increased by using an image-based screening system. This study aim to determine the efficacy of this screening AI algorithm, a prototype, in addition to or instead of screening of dry eye using a simple DEQ-5 symptom questionnaire.
Detailed description
Investigators have shown that a single corneal picture after dye staining can detect DED that are ideally managed at tertiary care because these require prescription eyedrops. The main type of DED patients that respond to cyclosporine eyedrops are those with severe cornea staining. In collaboration with data scientists from ASTAR, the preliminary data involving more than 1000 images from China and Singapore show that this artificial intelligence-based screening is sensitive and specific. By reducing unnecessary referrals to hospitals, investigators will make healthcare more sustainable and affordable. Previously, patients in the community are evaluated purely based on subjective symptoms. investigators not only standardize this with a validated and short DEQ5 questionnaire, but evaluate the accuracy of screening is improved by using the AI algorithms on the corneal image, a prototype, in addition to the DEQ5, and in place of the DEQ5. Aim: Determine the efficacy of this screening AI algorithm, a prototype, in addition to or instead of screening of dry eye using a simple DEQ-5 symptom questionnaire. Rationale: DEQ-5 is aimed to detect dry eye cases, but not necessarily dry eye requiring specialist care. The AI algorithm picks up cases with central cornea staining, which can then be referred for specialist care. Non-referred cases can be managed with eyelid warming, artificial tears and advice, with the aim of rescreening at a later time.
Interventions
None listed
Sponsors
Singapore National Eye Centre
Study design
Observational model
COHORT
Time perspective
CROSS_SECTIONAL
Eligibility
Sex/Gender
ALL
Age
21 Years to 99 Years
Healthy volunteers
Yes
Inclusion criteria
1. 21 years old and above 2. Participants must be previously diagnosed with dry eye in the dry eye clinic (previous referred and had various forms of treatment such as artificial tears or prescription eyedrops) 3. Willing to perform all eye examinations and questionnaires in this study 4. Ability to provide informed consent
Exclusion criteria
1. All subjects meeting any of the
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Determine the efficacy of this screening AI algorithm, a prototype, in addition to or instead of screening of dry eye using a simple DEQ-5 symptom questionnaire. | 3 years | DEQ-5 is aimed to detect dry eye cases, but not necessarily dry eye requiring specialist care. The AI algorithm picks up cases with central cornea staining, which can then be referred for specialist care. Non-referred cases can be managed with eyelid warming, artificial tears and advice, with the aim of rescreening at a later time. |
Countries
Singapore
Contacts
CONTACTSharon Yeo, BSc
PRINCIPAL_INVESTIGATORLouis Tong
Singapore Eye Research Institute (SERI)
Outcome results
None listed